Department of Chemistry, Virginia Tech, Blacksburg, VA 24061, USA.
J Chromatogr A. 2012 Mar 16;1229:237-48. doi: 10.1016/j.chroma.2011.12.089. Epub 2012 Jan 13.
A thorough evaluation of 5 μm bare silica from two major vendors for achiral supercritical fluid chromatography of polar analytes has been carried out. Columns were the same dimension, and a virgin column was reserved for each modifier-mixture combination. Three mixtures were prepared and chromatographically separated via a gradient of methanol-modified CO(2) that incorporated 5% (w/w) water as a neutral additive. Mixture (A) invoked both trifluoroacetic acid and water as additives. Mixture (B) utilized isopropyl amine and water; while mixture (C) employed either ammonium acetate and water as additives or only water. Regardless of the mixture components and mobile phase composition, duplicate separations with superior selectivity and excellent peak resolution were observed on five analysis days over a 15-day period. Subsequent removal of water (i.e. primary additive) from each of the mobile phases led to lower selectivity for early eluting components but excellent peak resolution prevailed for later eluting peaks during a later 5-day testing period with only the secondary additive. The re-introduction of 5% water into the mobile phase (after allowing the bare silica columns that were used with no water to sit for 30 days) slowly yielded the original separation after approximately five injections. A hydrophilic interaction liquid chromatography (HILIC)-like mechanism for SFC whereby analyte partitions between water absorbed on the silica and water in the mobile phase is proposed. The general utility of this experimental approach with bare silica was subsequently demonstrated by single injection of ten drug-like compounds with each of the four mobile phases that previously were utilized with the three model compound mixtures. In each case, sharp peaks were observed for each drug-like compound regardless of the additive although retention times varied with the additive employed.
对两个主要供应商的 5 μm 裸硅胶进行了彻底评估，用于手性超临界流体色谱法分离极性分析物。柱的尺寸相同，为每种改性剂-混合物组合保留了一根新柱。制备了三种混合物，并通过包含 5%（w/w）水作为中性添加剂的甲醇改性 CO2 梯度进行色谱分离。混合物（A）同时使用三氟乙酸和水作为添加剂。混合物（B）使用异丙胺和水；而混合物（C）则使用乙酸铵和水或仅水作为添加剂。无论混合物成分和流动相组成如何，在 15 天的时间内进行了 5 天的重复分离，均观察到选择性更好、峰分离度更高。随后，从每个流动相中除去水（即主要添加剂），但对于后期洗脱的峰，在仅使用次要添加剂的后续 5 天测试期内，仍保持良好的峰分离度。在将无水移动相重新引入流动相（在允许使用无水的裸硅胶柱放置 30 天后）后，大约经过五次进样后，缓慢恢复了原始分离。提出了一种超临界流体色谱中亲水相互作用的机制，即分析物在硅胶上吸收的水和流动相中的水之间分配。随后，用这四种流动相中的每一种对十个类似药物的化合物进行单次进样，证明了裸硅胶的这种实验方法的通用性，这些流动相之前曾与三种模型化合物混合物一起使用。在每种情况下，尽管保留时间随所使用的添加剂而变化，但每个类似药物的化合物都观察到尖锐的峰，无论添加剂如何。
